Structural and magnetic studies on macrocyclic dicopper (II

Contribution from the Departmentof Chemistry, Memorial University of Newfoundland, St. John's, Newfoundland, Canada A1B 3X7,. Division of Chemistry ...
0 downloads 0 Views 520KB Size
1324

Inorg. Chem. 1990, 29, 1324-1327

Contribution from the Department of Chemistry, Memorial University of Newfoundland, St. John's, Newfoundland, Canada A I B 3 x 7 , Division of Chemistry, National Research Council, Ottawa, Ontario, Canada K I A OR6, and Department of Inorganic Chemistry, Indian Association for the Cultivation of Science, Calcutta 700032, India

Structural and Magnetic Studies on Macrocyclic Dicopper(I1) Complexes. Influence of Electron-Withdrawing Axial Ligands on Spin Exchange+ Sanat K. Mandal,Ia Laurence K. Thompson,*?laMichael J. Newlands,Ia Eric J. Gabe,Ib and Kamalaksha Nag1C Received June 30, 1989 The antiferromagnetically coupled macrocyclic binuclear copper(I1) complexes [Cu2(UPM)X2].yH20(X = CI, Br, I, N3) (UPM derived by template condensation of 4-methyl-2,6-diformylphenolwith 1,3-diaminopropane) are structurally similar with a trans-axial arrangement of ligands X. Exchange integrals fall in the range -2J = 720-855 cm-I with CI < Br < N, < I. The com lex [Cu2(UPM)Br2].4H20(11) crystallizes in the monoclinic system, space group C2/m,with a = 7.9196 (13) A, b = 17.1380 (4) , c = 11.2232 (14) A, p = 92.1 11 (13)O, and Z = 2. The complex [Cu2(UPM)12].H20(111) crystallizes in the monoclinic system, space group P2,/a, with a = 8.2826 (8) A, b = 16.7383 (8) A, c = 9.6470 (12) A, p = 99.107 (14)O, and Z = 4. Both complexes have an essentially flat, macrocyclic, binuclear structure involving square-pyramidal copper(I1) centers. The halogens are bound in a trans-axial arrangement (Cu-Br = 2.737 (2) 8,;Cu-I = 3.026 (2) A, and the copper centers are displaced toward the halogen from the basal N 2 0 2donor set (Br, 0.205 (4) 8,;I, 0.176 (5) A).

8:

Introduction

Macrocyclic dicopper( 11) complexes derived from 2,6-diformyland 2,6-diacetyl(4-substituted)phenols (Figure 1) a r e numerous a n d have been studied from t h e standpoint of their interesting structural, magnetic, EPR, and electrochemical properties a n d also a s potential models for binuclear copper metalloprotein active sites.2-18 In general, systems of this sort are strongly antiferromagnetically coupled and exhibit one-electron reduction steps a t negative potential^.^,^,'^^^,'^,'^^'^ In a few cases involving ligands with saturated nitrogen donor groups, one- and even two-electron oxidation steps can be observed a t positive potentials (1-1.5 V vs SCE).14,'5 T h e reduction potentials can be changed by t h e inclusion of electron-withdrawing or electron-donating groups (R and R 1in Figure l)'o,'8 and by varying the size of the macrocyclic ring (R2 in Figure Electron-withdrawing groups (e.g. R = CF3vs C H 3 ) make both one-electron reduction potentials more positive,'8 while electron-releasing groups ( e g R 1 = CH3 and CH3CH2CH2vs H) seem only to influence the second one-electron reduction step, making it more negative (vs SCE).'O I n this study we have examined a series of compounds [Cu2( U P M ) X 2 ] - y H 2 0 ( X = CI,Br, I, N3) (Figure 1, R = C H 3 , RI = H , R2 = ( C H 2 ) 4 from t h e standpoint of their structural a n d magnetic properties. T h e complex [Cu2(UPM)CI2].6H20(I) has already been reported2 and consists of a binuclear phenoxo-bridged macrocyclic structure5 involving one terminally bound chlorine per copper in a trans-axial arrangement. T h e structures of the bromo a n d iodo analogues, [ C u 2 ( U P M ) B r 2 ] . 4 H 2 0 (11) a n d [Cu2(UPM)12].H20 (III), a r e very similar with the halogens again bound in a trans-axial arrangement. The azide derivative, [ C U ~ ( U P M ) ( N ~ ) ~ ] . ~(VI), . ~ H ~is Oassumed to have a similar structure, on t h e basis of spectroscopic evidence. All of the complexes exhibit strong, net antiferromagnetism with -2J falling in the range 720-855 cm-I, and the differences in exchange can be attributed to electronic effects associated with the halogen and azide ligands. l).11919

Experimental Section

Table I. Crystal Data for [Cu2(UPM)Br2].4H20(11) and [ C U ~ ( U P M ) I ~ ] . H(111) ~O (a) Compound I1 chem formula: fw: 759.44 C U ~ C ~ ~ H ~ ~ B ~ space ~ Ngroup: ~ ~ C2/m ~ - ~ H ~ ~ a = 7.9196 (13) A T = 295 K b = 17.1380 (4) 8, X = 0.70930 8, c = 11.2232 (14) 8, paid = 1.66 g 1 = 4.03 mm-' /3 = 92.111 (13)' V = 1522.25 A3 R = 0.039 z=2 R, = 0.042 (b) Compound I11 chem formula: CUZC24H241

2N402'H20

= 8.2826 (8) 8, b = 16.7383 (8) 8, c = 9.6470 (12) 8, /3 = 99.107 (14)' V = 1320.57 A' 2 = 4 a

fw: 399.70 space group: P2'/a T = 295 K X = 0.709 30 8, paid = 2.01 g cm-) p = 3.97 mm-' R = 0.053 R, = 0.052

filtered off, washed with a water/methanol mixture, and recrystallized from acetonitrile to give dark greenish brown crystals (yield 0.43 g). ~ H ~were O [Cu2(UPM)Br2].4H20(11) and [ C U ~ ( U P M ) ( N , ) ~ ] . ~ .(VI) prepared in a similar manner by using sodium bromide and sodium azide, respectively. [Cu2(UPM)C12].6H20(I) was prepared according to the ( 1 ) (a) Memorial University. (b) National Research Council. (c) Indian

Association for the Cultivation of Science. (2) Pilkington, N. H.; Robson, R. Aust. J . Chem. 1970, 23, 2225. (3) Okawa, H.; Kida, S. Bull. Chem. SOC.Jpn. 1972, 45, 1759. (4) Addison, A. W. Inorg. Nucl. Chem. Lett. 1976, 12, 899. (5) Hoskins, B. F.; McLeod N. J.; Schaap, H. A. Aust. J . Chem. 1976,29, 515.

(6) Gagnt, R. R.; Koval, C. A,; Smith, T. J. J . Am. Chem. SOC.1977, 99, 8367. (7) Lambert, S. L.; Hendrickson, D. N. Inorg. Chem. 1979, 18, 2683. (8) Gagnt, R. R.; Koval, C. A,; Smith, T.J.; Cimolino, M. C. J . Am. Chem. SOC.1979, 101, 4511. (9) Gagnt, R. R.;Henling, L. M.; Kistenmacher, T. J. Inorg. Chem. 1980, 19, 1226. (10) Mandal, S. K.; Nag, K. J . Chem. SOC.,Dalton Trans. 1983, 2429. ( 1 1 ) Long, R. S.; Hendrickson, D. N . J . Am. Chem. SOC.1983, 105, 1513. (12) Mandal, S. K.; Nag, K. J . Chem. Soc., Dalton Trans. 1984, 2141. ( 1 3) Carlisle, W. D.; Fenton, D. E.; Roberts, P. B.; Casellato, U.;Vigato, P. A,; Graziani, R. Transition M e t . Chem. 1986, 11, 292. (14) Mandal, S. K.; Thompson, L. K.; Nag, K.; Charland, J.-P.;Gabe, E.

Physical Measurements. Variable-temperature magnetic susceptibility data were obtained in the temperature range 5-300 K by using an Oxford Instruments superconducting Faraday magnetic susceptibility system with a Sartorius 4432 microbalance. A main solenoid field of 1.5 T and a gradient field of I O T m-I were employed. HgCo(NCS)4 was used as calibrant. Infrared spectra were obtained with a Perkin-Elmer 283 J . Inorg. Chem. 1987, 26, 1391. spectrometer. (15) Mandal, S. K.; Thompson, L. K.; Nag, K.; Charland, J.-P.; Gabe, E. Synthesis of Complexes. [Cu2(UPM)12].H20(111). An aqueous soJ. Can. J . Chem. 1987, 65, 2815. lution ( I O mL) of sodium iodide (0.40 g, 2.7 mmol) was added to a (16) Mandal, S. K.; Thompson, L. K.; Nag, K. Inorg. Chim. Acta 1988, 149, stirred, hot, aqueous solution (100 mL) of [ C U ~ ( U P M ) ( C I O ~ ) ~ ] ~ ~ H ~ O ~247. (0.50 g, 0.66 mmol). The resulting brown, microcrystalline product was ( 1 7 ) Lacroix, P.; Kahn, 0.; Gleizes, A,; Valade, L.; Cassoux, P. Nouc. J . Chim. 1984, 8, 643. (18) Lacroix, P.; Kahn, 0.;Theobald, F.; Leroy, J.; Wakselman, C. Inorg. Chim. Acra 1988, 142, 129. *To whom correspondence should be addressed (19) Mandal, S. K.; Thompson, L. K.; Newlands, M. J.; Gabe, E. J. Inorg. ' NRCC Contribution No. 30799. Chem. 1989, 28, 3707.

0020-1669/90/1329-1324$02.50/0

0 1990 American Chemical Society

Inorganic Chemistry, Vol. 29, No. 7, 1990 1325

Structural and Magnetic Studies on Cu1Il Complexes

R

v I

R Figure 1. Macrocyclic ligands: UPM (R = CH,, R1= H, R2 = (CH2)J; UBM (R = CH,, R1 = H, R2 = (CH2)4). Table 11. Final Atomic Positional Parameters and Equivalent Isotropic Debye-Waller Temperature Factors (Esd's) for [Cu2(UPM)Br2].4H20 ( I I ) Y

X

Br Cu

0.26328 (18) 0.56442 (17)

0 N CI

0.6606 (8)

C2 C3 c4

0.4322 (9) 0.4350 (9)

C6

0.6449 (IO) 0.7509 (17) 0.731 (3) 0 0.7599 (8)

cs

C7 C8 OH1 OH2

I/* I12

0.0718 0.0862 0.1482 0.1920 0.2742 0.3154 0.4048 0.1582 0.0717 0 0.2458 0.3253

B;.,,"

Z

0 0

0.24241 (12) 0.37006 ( 1 1) (4) (4) (5) (4) (4) (6) (7) (4) (5) (6) (4)

0.2779 (5) '12

0.5948 (6) 0.5924 (6) '/2

0.3040 0.1645 0.1162 0 0.1292

(6) (9) (13) (6)

A'

5.34 (8) 2.65 (6) 2.7 (3) 3.7 (3) 2.5 (4) 2.6 (3) 2.9 (3) 3.2 (5) 4.3 (6) 3.3 (3) 8.1 (7) 9.3 (12) 6.75 (23) 6.43 (16)

Bi, is the mean of the principal axes of the thermal ellipsoid. literature procedure.2 C, H, N, and halogen analyses were carried out by Canadian Microanalytical Service, Vancouver, Canada, and all compounds were found to be analytically pure. Crystallographic Data Collection and Refinement of the Structures. [Cu2(UPM)Br2].4H20(11). Crystals of I1 are green in color. The diffraction intensities of an approximately 0.10 x 0.15 X 0.20 mm crystal were collected with graphite-monochromatized Mo Ka radiation by using = 45.0' on an the 8/28 scan technique with profile analysis20 to ,2,8, Enraf-Nonius CAD4 diffractometer at 295 K. A total of 2722 reflections were measured, of which 1034 were unique and 752 were considered significant with I,,, > 2.5a(fn,). Lorentz and polarization factors were applied, but no correction was made for absorption. The cell parameters were obtained by the least-squares refinement of the setting angles of 18 reflections with 28 = 40-45' (X(Mo Ka) = 0.70930 A). The structure was solved by direct methods using M U L T A N ~ 'and refined by full-matrix least-squares methods to tinal residuals of R = 0.039 and R , = 0.042 for the significant data (0.067 and 0.067 for all data) and unit weights. Crystal data are given in Table I, and final atomic positional parameters and equivalent isotropic temperature factors are listed in Table 11. All calculations were performed with the NRCVAX system of programs,22and scattering factors were taken from ref 23. Anisotropic thermal parameters (Table SI) and a listing of structure factors are included as supplementary material. [Cu2(UPM)I2].H2O(111). A greenish brown crystal of Ill,of dimensions 0.20 X 0.20 X 0.20 mm, was examined crystallographically in the same manner as 11. A total of 6845 reflections were measured, of which (20) Grant, D. F.;Gabe, E. J . J . Appl. Crystallogr. 1978, 11, 114. ( 2 1 ) Germain, G.; Main, P.; Woolfson, M . M. Acta Crystallogr. Sect. A 1971, A27. 368. (22) Gabe, E. J.; Lee, F. L.; LePage, Y . In Crystallographic Computing III;

Sheldrick, G.,Kruger, C., Gcddard, R., Eds.; Clarendon Press: Oxford, England, 1985; p 167. (23) International Tables for X-ray Crystallography; Kynoch Press: Birmingham, England, 1974; Vol. IV, Table 2.28, p 99.

Figure 2. Structural representation of [Cu2(UPM)Br2](11) with hydrogen atoms omitted (40% probability thermal ellipsoids). Table 111. Final Atomic Positional Parameters and Equivalent Isotropic Debye-Waller Temperature Factors (Eds's) for [ C U ~ ( U P M ) I ~ ] * (111) H~O X Y Z Bi,? A2 I 0.26496 (9) 0.09460 (5) 0.77153 (9) 4.14 (3) 0.93259 (14) 0.04297 (7) 0.62146 (12) 2.42 (4) Cu 0.9946 (8) 0.0628 (3) 0.4352 (6) 2.50 (24) 0 N1 0.8530 (10) 0.0011 (5) 0.7889 (8) 2.8 (3) N2 0.8287 (9) 0.1479 (5) 0.6207 (8) 2.8 (3) 1.0105 ( I O ) 0.1322 (6) 0.3744 (9) 2.5 (3) C1 C2 1.0822 (11) 0.1394 (6) 0.2489 (9) 2.6 (3) C3 C4 C5 C6 C7

C8 C9

C10 C11 C12

1.0966 (12) 1.0485 (12) 1.0712 (21) 0.9752 (12) 0.9546 (11) 0.8641 (11) 0.7238 (16) 0.7999 (17) 0.7815 (18) 0.8597 (13)

0.2149 0.2852 0.3660 0.2777 0.2037 0.2059 0.1668 0.1400 0.0517 -0.0727

(6) (6) (8) (6) (5) (5) (7) (7) (7) (6)

0.1887 0.2474 0.1852 0.3657 0.4305 0.5429 0.7248 0.8683 0.8922 0.8247

(11)

(IO) (16)

(IO) (9)

(IO) (14) (11) (13) (11)

3.0 3.0 4.5 2.9 2.5 2.5 3.9 3.9 4.1 3.2

(4) (4) (6) (4) (3) (3) (5) (5) (5) (4)

Bisois the mean of the principal axes of the thermal ellipsoid. Table IV. Interatomic Distances (A) and Angles (deg) Relevant to the Copper Coordination Spheres in [Cu2(UPM)Br2].4H20(11) Br-Cu 2.7365 (20) CU-N 1.972 (6) cu-0 1.989 (4) CU-NB 1.972 (6) CU-OA 1.989 (4) 0-CUA 1.989 (4) CU-CUA 3.124 (2)

Br-Cu-0 Br-Cu-OA Br-Cu-N Br-Cu-NB 0-CU-OA 0-CU-N

98.15 98.15 94.1 1 94.11 76.43 92.04

(5) (5) (20) (20) (20) (22)

0-CU-NB OA-CU-N OA-CU-NB N-Cu-Nb CU-0-CUA

164.21 (21) 164.21 (21) 92.04 (22) 97.0 (3) 103.6 (3)

3842 were unique and 2174 were considered significant with I,,, > 2.5u(Ine,). Crystal data are given in Table I, and final atomic positional

parameters and equivalent isotropic temperature factors are listed in Table 111. Anisotropic thermal parameters (Table SII) and a listing of structure factors are included as supplementary material.

Results and Discussion Description of the Structures of [Cu2(UPM)Br21.4H20(11) and [Cu2(UPM)12].H20(111). T h e structure of I1 is shown in Figure 2 , and interatomic distances a n d angles relevant to t h e copper coordination spheres a r e given in Table IV. T h e macrocyclic ligand is almost flat, enclosing two distorted square-pyramidal copper( 11) centers, bridged by phenoxide oxygen atoms and bound terminally in a trans arrangement to two bromine atoms. T h e copper-bromine separation is quite long ( 2 . 7 3 7 (2) A), a n d the copper center is displaced from the strictly planar N 2 0 2donor set toward the bromine by 0.205 (4) A. T h e copper-copper distance is 3.124 (2) A, and the phenoxide bridge angle (Cu-0C u ) is 103.6 ( 3 ) ' . Structurally, 11 is very similar to the chloro-

1326 Inorganic Chemistry, Vol. 29, No. 7, 1990

Mandal et al. 4.0

I I

C6

-0 0

50

100

150

200

250

300

TEMP ( K l

'-pl" Figure 3. Structural representation of [Cu2(UPM)12](111) with hydrogen atoms omitted (40% probability thermal ellipsoids). Table V. Interatomic Distances (A) and Angles (deg) Relevant to the Copper Coordination Spheres in [Cu2(UPM)12].H20(111) I-CUA 3.0264 ( 1 5 ) CU-N 1 1.968 (8) CU-18 3.0264 ( 1 5 ) CU-N~ 1.956 (8) cu-0 1.974 (6) 0-CUC 1.975 (6) CU-OC 1.975 (6) CU-CUA 3.104 (2)

IB-CU-0 IB-CU-OC IB-CU-NI IB-CU-N~ 0-CU-OC

92.27 (19) 95.38 (19) 96.14 (24) 96.41 (24) 76.34 (24)

0-CU-N! O-CU-N~ OC-CU-NI OC-CU-N~ NI-CU-N~

167.5 (3) 91.4 (3) 93.7 (3) 163.3 (3) 96.8 (3)

analogue [Cu2(UPM)C12]-6H20,5which has a copper-copper separation of 3.133 A and a phenoxide bridge angle of 104.5' and square-pyramidal copper centers with similar apical displacements. The solid angle at the phenoxide bridge in I1 is exactly 360.0', whereas in [Cu2(UPM)C12].6H20an angle of 357.4' indicates slight pyramidal distortion. The structure of 111 is shown in Figure 3, and interatomic distances and angles relevant to the copper coordination spheres are given in Table V. Structurally, 111 is very similar to the bromo analogue (11) with a trans square-pyramidal macrocyclic structure. Copper-nitrogen and copper-oxygen distances are somewhat shorter than those in 11, but the phenoxide bridge angle (103.7') is essentially the same. The copper-copper separation (3.104 A) is slightly shorter, in keeping with reduced dimensions within the CuZO2framework. The N 2 0 zdonor set is almost planar, and the copper centers are displaced from their mean plane by 0.176 (5) A toward the distant iodine atoms (Cu-I = 3.026 (2) A). The phenoxide bridge has almost trigonal-planar character with a solid angle of 359.1". Variable-temperature magnetic susceptibility measurements were performed on powdered samples of compounds 1-111 in the temperature range 5-300 K. The results are summarized in Table VI. The best fit of the data to the modified Van Vleck equationz4 for exchange-coupled pairs of copper(l1) ions (eq 1) was deterXU

=

mined in each case with a two-variable nonlinear regression (24) Van Vleck, J. H. The Theory of Electric and Magnetic Susceptibilities; 1932, Oxford University Press: London, 1932; Chapter 9.

Figure 4. Magnetic susceptibility data for [Cu2(UPM)Br2].H20(11). The solid line was calculated from eq 1 with g = 2.06 ( I ) and -2J = 771 (4) cm-' (corrected for 0.38% paramagnetic impurity).

analysis.z5 In this expression -2J (in the spin Hamiltonian H = -2Ji1.i2) is the singlet-triplet splitting or exchange integral and other terms have their usual meaning. p represents the fraction of a possible magnetically dilute, mononuclear copper(I1) impurity. The temperature-independent paramagnetism Na, was taken as cgs units/mol copper, and p was treated as a floating 60 x parameter. Strong net antiferromagnetic exchange was observed in all cases, with -2J falling in the range 720-855 cm-I. A typical experimental variable-temperature magnetic susceptibility data plot, including the best fit theoretical line, for compound I1 is shown in Figure 4. Similar plots for the other compounds are available as supplementary material. A comparison of key structural parameters and exchange integrals for 1-111 (Table VI) reveals that despite quite similar binuclear center dimensions, e.g. Cu-0-Cu bridge angles, CuO(phenoxide) separations, Cu-Cu separations, out of plane copper displacements, and the solid angles at the phenoxide bridge, exchange integrals vary quite dramatically with CI (722 cm-I) < Br (771 cm-I) < I (852 cm-I). The Cu-0-Cu bridge angle for I is the largest of the group, and on the basis of the primary importance of this feature to superexchange, the low value of the exchange integral for I is quite significant. The exchange integral for I has been reported in a previous study (-2J = 588 cm-')' and differs significantly from the value we have observed. The trend of diminished exchange for the chloro complex is, however, still apparent. A comparison of the structural and magnetic features of 111 with those of the perchlorate complex V, which involves two different molecules in the unit cell, one involving six-coordinate and the other five-coordinate copper(1I) centers, reveals similar binuclear center dimensions, except for the more pronounced displacement of the copper centers from the basal plane in 111, and comparable exchange. A further comparison of iV and V (Figure 1; UBM, R = Me, R' = H, Rz = (CHJ4), which involves six-coordinate copper(I1) centers, shows that minor perturbations to the binuclear center dimensions, including significant, but small, pyramidal distortion at the phenoxide bridge, does not affect exchange ~ignificant1y.l~ Pyramidal distortions at oxygen bridges in systems of this sort have been shown to influence exchange coupling, but usually much more severe distortion than observed in e.g. V is required to significantly influence the exchange proc e s ~ . ~ ~u ?overlap ~' between the copper magnetic orbitals and the oxygen bridge is likely to be reduced to some extent as a result of out of plane distortion at the copper center, but for the small distortions apparent in these complexes the effect is likely to be ( 2 5 ) Duggleby, R. G. Anal. Biochem. 1981, 110, 9. (26) Murry, K. S. In Copper Coordination Chemistry: Inorganic and Biological Perspectives; Karlin, K. D., Zubieta, J., Eds.; Adenine Press: New York, 1985. (27) Mazurek, W.; Kennedy, 8. J.; Murray, K. S.; OConnor, M. J.; Rogers, J. R.; Snow, M. R.; Wedd, A. G.;Zwack, R. R. Inorg. Chem. 1985.24, 3258.

Inorganic Chemistry, Vol. 29, No. 7, 1990 1327

Structural and Magnetic Studies on CuI12 Complexes

Table VI. Key Structural Data and Magnetic Data for Macrocyclic Dicopper(l1) Complexes

complexa [CU~(UPM)CI~].I SH2O ( I ) [Cu2(UPM)Br2].H20(11) [ C U ~ ( U P M ) I ~ ] *( H I l l~) O

Cu-0-Cu, deg

ZO, deg

Cu-Cu, A

Cu-0, A 1.981 (2) 1.989 (4) 1.974 (6) 1.975 (6) 1.981 (6) 1.989 (6) 1.943 (2) 1.970 (2)

104.5 ( I ) 103.6 (3) 103.7 (3)

357.4 360.0 359.1

3.133 (1) 3.124 (2) 3.104 (2)

[ C U ~ ( U P M ) ( C I O ~ ) ~ ](.I~VH) ~ O 102.3 (3)

358.4 359.1 353.2

3.091 (3) 3.096 (3) 3.0354 (7)

[Cu,(UBM)(CIO4)21 (V)

103.6 (3) 101.77 (9)

[ C U ~ ( U P M ) ( N J ~ I *(VI) H~O a

out-of-plane displacement, A 0.21 0.205 (4) 0.176 (5) 0.019 (4) 0.083 (6) 0.004 ( I )

-23, 722 771 852

cm-' (20) (4) (9)

g 2.12 (5) 2.06 ( I ) 2.13 (4)

850 (2)b 857 (6)b 780 (7)

2.15 (4)

Formulas for 1, 11, and V I differ from those reported earlier due to loss of solvent on drying prior to magnetic measurements. bReference 19.

small,28 so that this geometric feature is not likely to dominate the exchange situation. The marked difference in exchange between compounds 1-111 and the -25 trend CI < Br < I, which parallels the trend in electronegativity of the halogens, clearly indicates that the axially bound halogens have a primary influence on exchange, even though they are bound orthogonally to the copper magnetic orbital. The similarity in the magnitude of the exchange integrals of compounds Ill-V can be rationalized in terms of a general similarity in binuclear center dimensions but also on the basis that iodine has a small electron-withdrawing effect and would not polarize electron density in the copper magnetic orbital manifold to a significant extent. The much smaller values of exchange integral observed for I and I 1 indicate that the more electronegative bromine and chlorine ligands are effectively polarizing the electron density in the copper magnetic orbitals, with the result that I and I1 exhibit more paramagnetism than 111. The effect can be imagined as a modulation of the molecular microcircuitry associated with superexchange between the two copper centers. The observation of such an orthogonal perturbation is unusual, but it has been observed before in cases where isostructural pairs of chloro and bromo complexes have involved terminal halogens bound directly to copper magnetic orbital^.^^-^^ Variable-temperature magnetic data for the azido complex (VI) (5-300 K; data treatment as for 1-111) are given in Table VI (a susceptibility data plot is available as supplementary material). This complex is strongly coupled with -2J = 780 (7) cm-I, which is between the values of compounds I and IV. Structurally, VI is assumed to be similar to compounds 1-111, with a trans arrangement of terminally bound azide groups. The infrared spectrum shows a very strong, single azide absorption at 201 5 cm-', which can be associated with a highly symmetric trans-bis(azido) derivative.,' The fact that spin exchange for the azido compound is comparable with that of the bromo derivative (11) suggests that, assuming a similarity in the binuclear center dimensions, the azide ligands are also exhibiting a polarizing effect on the electron density in the copper magnetic orbital envelope. The copper magnetic orbitals interact directly with the phenoxide bridging oxygen atoms in this type of complex, and so their delocalization should be affected by the electronegativity of the bridging atom and possibly by perturbations within the phenoxide (28) Kahn, 0. Angew. Chem., Int. Ed. Engl. 1985, 24, 834. (29) Thompson, L. K.; Mandal, S.K.; Charland, J.-P.;Gabe, E. J. Can. J . Chem. 1988, 66, 348. (30) Thompson, L. K.; Lee, F. L.; Gabe, E. J. Inorg. Chem. 1988, 27, 39. (31) Nakamoto. K. Infrared and Raman Spectra of Inorganic and Coordination Compounds, 3rd ed.; John Wiley and Sons: New York, 1978.

group itself. In a study of the complexes [CU~(L)](CIO,)~ (Figure 1; L = UPM and UPMCF,, R = CH3 and CF,, respectively), which have comparable binuclear, macrocyclic structures,18 the magnetic properties are essentially the same (-2J = 710 and 712 cm-I, respectively), indicating that the electron-withdrawing CF3 group does not influence magnetic exchange significantly, despite the fact that metal-centered electrochemical reductions are shifted to more positive potentials in the CF, derivative. However, in other binuclear copper(11) complexes involving the bridging ligands oxalate, oxamide, and dithioxamide, which have related structures, the effect of reduced electronegativity of the bridging atoms is clearly reflected in a trend of increasing antiferromagnetic exchange resulting from increased delocalization of the copper magnetic orbitals.28 One way of testing this effect in a macrocyclic system of the type under discussion would be to replace the phenoxide oxygen atom directly with sulfur. This has, to our knowledge, not yet been done. However, in a related, nonmacrovclic system a comparison of pentadentate, binucleating phenoxide and thiophenoxide bridging ligands derived from 4-methyl-2,6-diformylphenol and 4-methyl-2,6-diformylthiophenolshows that the phenoxide-bridged dicopper(I1) derivatives are strongly antiferromagnetically coupled but the thiophenoxide analogues unexpectedly exhibit almost Curie-like behavior.32 The significant reduction in antiferromagnetic exchange in the sulfur derivatives is attributed to the marked pyramidal distortion (sp3 character) of the sulfur bridge, which is not the case for the analogous oxygen-bridged derivatives, which have mainly sp2 character. In the present study a very different perturbation on the copper magnetic orbitals has been demonstrated, and it indicates that fine tuning of the magnetic exchange in binuclear copper(I1) complexes can be effected in subtle ways that do not involve a direct interaction with those magnetic orbitals. Acknowledgment. We thank the Natural Sciences and Engineering Research Council of Canada for financial support for this study. Supplementary Material Available: Anisotropic thermal parameters and bond length and angle data for I1 (Tables SI and SIII, respectively) and 111 (Tables SI1 and SIV, respectively), atomic positional parameters for hydrogen atoms in I1 and 111 (Tables SV and SVI, respectively), and variable-temperature magnetic susceptibility plots for I, 111, and VI (Figures S 1-S3, respectively) (12 pages); observed and calculated structure factors for I1 and 111 (69 pages). Ordering information is given on any current masthead page. (32) Iliopoulos, P.; Murray, K. S.; Robson, R.; Wilson, J.; Williams, G. A. J . Chem. Soc., Dalton Trans. 1987, 1585.